So instead of a hump at 12:30, they've created two humps at 11:30 and 1:30?
Why not offer a continuous discount curve? If you show up at 1:25, you get 23% off your meal, so there is only a 2% discount for dawdling 5 extra minutes in the food line before paying.
I love how some companies are "smart" but get every details wrong anyways.
When I was in Argentina in 2003, there were giant internet cafés (up to 250 computers I believe) were the price of the minute of internet was floating depending on how many people were online: you had a max price, a min price and the current price was just the proportion of used computers (so it was not a bidding market). You had a few screens on the wall telling you the current price (I think they billed at the beginning of each minute).
In the middle of the night the price could get quite low.
Could do, but why tie to time? What you actually trying to minimise is time spent queueing. So record when people arrive and apply a discount/surcharge depending on how long they queued (or just based on the number of people ahead of them in the queue).
Seemingly perversely, a long wait would imply a higher cost for the meal.
It's basically congestion charging based on queue length/time rather than time of day.
The main advantage of tying it to time is that you can let people plan for it. If you have dynamic pricing based on unpredictable conditions (how many people are in the cafeteria right now?) you can't plan for it, and the best case scenario is that employees get up, go to the cafeteria, see that the demand/price is too high at the moment, go back to their desk, come back 10 minutes later, repeat until the price/demand is low enough.
This was my thought too, hand out a set of discount cards (or passbook barcodes) with a moving window from 11:30 to 1:30 on them. If someone doesn't like their time frame, they could swap it with a different employee.
I don't like pre-planned organization, it adds cruft. Maybe just do a "current discount" depending on the people in the line, and put it on a mobile website ?
Even that relies on a model of the utilization of the cafeteria over time. If Goldman wanted to incentivize spending a minimal amount of time in the cafeteria, they should just measure that. For example, offer a discount of (25 - X)% where X is the number of minutes between entering the cafeteria and paying for your food. People would learn to anticipate events such as backed-up food or cash register lines and adjust their behaviour appropriately.
Unless someone has observed otherwise, I'm betting continuous wouldn't work. The difference between "now" and "later" won't be as big, and it will take effort to calculate, which makes it feel smaller too. To effect people's day-to-day habits, you need to present a motivator to the affective decision centers.
A continuous discount curve would be hard to implement and hard to explain.
The two humps will be smaller than the single big hump.
The 11:30 hump would be a race to beat the cutoff.
The 13:20 hump is the delaying described in the article.
And there's probably still a hump at 12:30, it just doesn't last an hour anymore.
I was thinking the same thing. Continuous incentives are better, to the extent that they're not too complicated. But it's Goldman Sachs - I'm sure they can figure it out.
Comments
So instead of a hump at 12:30, they've created two humps at 11:30 and 1:30?
Why not offer a continuous discount curve? If you show up at 1:25, you get 23% off your meal, so there is only a 2% discount for dawdling 5 extra minutes in the food line before paying.
I love how some companies are "smart" but get every details wrong anyways. When I was in Argentina in 2003, there were giant internet cafés (up to 250 computers I believe) were the price of the minute of internet was floating depending on how many people were online: you had a max price, a min price and the current price was just the proportion of used computers (so it was not a bidding market). You had a few screens on the wall telling you the current price (I think they billed at the beginning of each minute). In the middle of the night the price could get quite low.
Could do, but why tie to time? What you actually trying to minimise is time spent queueing. So record when people arrive and apply a discount/surcharge depending on how long they queued (or just based on the number of people ahead of them in the queue).
Seemingly perversely, a long wait would imply a higher cost for the meal.
It's basically congestion charging based on queue length/time rather than time of day.
The main advantage of tying it to time is that you can let people plan for it. If you have dynamic pricing based on unpredictable conditions (how many people are in the cafeteria right now?) you can't plan for it, and the best case scenario is that employees get up, go to the cafeteria, see that the demand/price is too high at the moment, go back to their desk, come back 10 minutes later, repeat until the price/demand is low enough.
pub/sub the queue time discount automatically from register transaction rate; client side compile local vendors and approximate arrival rate by gps.
This was my thought too, hand out a set of discount cards (or passbook barcodes) with a moving window from 11:30 to 1:30 on them. If someone doesn't like their time frame, they could swap it with a different employee.
...there's almost an app in there somewhere.
I don't like pre-planned organization, it adds cruft. Maybe just do a "current discount" depending on the people in the line, and put it on a mobile website ?
Sorry to be offtopic, but is your nick a typo? The german "saalwächter" would translate to "saalwaechter".
Yes, but an old typo. Name got misspelled about a hundred years back or so.
Even that relies on a model of the utilization of the cafeteria over time. If Goldman wanted to incentivize spending a minimal amount of time in the cafeteria, they should just measure that. For example, offer a discount of (25 - X)% where X is the number of minutes between entering the cafeteria and paying for your food. People would learn to anticipate events such as backed-up food or cash register lines and adjust their behaviour appropriately.
Unless someone has observed otherwise, I'm betting continuous wouldn't work. The difference between "now" and "later" won't be as big, and it will take effort to calculate, which makes it feel smaller too. To effect people's day-to-day habits, you need to present a motivator to the affective decision centers.
A continuous discount curve would be hard to implement and hard to explain. The two humps will be smaller than the single big hump. The 11:30 hump would be a race to beat the cutoff. The 13:20 hump is the delaying described in the article. And there's probably still a hump at 12:30, it just doesn't last an hour anymore.
I was thinking the same thing. Continuous incentives are better, to the extent that they're not too complicated. But it's Goldman Sachs - I'm sure they can figure it out.